4097 — All-Cause Mortality associated with Opioid and Benzodiazepine Co-Prescribing among Patients with PTSD

Lead/Presenter: Eric Hawkins, COIN - Seattle/Denver
All Authors: Hawkins EJ (Coin - Seattle/Denver) Malte CA (Coin - Seattle/Denver) Goldberg SB (VA Puget Sound HCS) Saxon AJ (Coin - Seattle/Denver)

Objectives:
The rise in pharmaceutical overdose deaths has disproportionately affected those with psychiatric disorders. Posttraumatic stress disorder (PTSD) is one of the most common psychiatric conditions treated in the Veteran Affairs (VA) Healthcare system. Veterans with PTSD present with multiple and complex symptoms, including anxiety, insomnia and pain, that impact quality of life and are often treated with opioid analgesic and benzodiazepine medications. However, overdose deaths underestimate the harm of these medications, as over-sedation causes fall-related injuries and motor vehicle accidents that contribute to mortality, accidents that may occur more often among concurrent users of these medications. Limited information exists on the comparative safety of opioids and benzodiazepines among patients with PTSD

Methods:
This propensity score-matched cohort analysis used data from the Corporate Warehouse Data (CDW) to identify Veterans with PTSD and VA Vital Status Files to assess mortality. Eligible participants received care at a VA facility and were diagnosed with PTSD during 2010-2011. The outcome was one-year all-cause mortality. Medication exposure was defined as new opioid only, benzodiazepine only and opioid and benzodiazepine concurrent therapies of > 30 days. Hazard rations (HR) from Cox proportional regressions estimated risk of all-cause mortality in patients newly co-prescribed opioids and benzodiazepines, as compared to propensity score-matched cohorts of opioids only, benzodiazepines only and non-users of these medications.

Results:
Incident rates of opioid only, benzodiazepine and concurrent use were 23.6, 14.1, and 1.72, respectively. The 12-month adjusted HR for death among concurrent users (n = 4,379, 115 deaths) was 1.45 (CI: 1.09-1.93) and 2.04 (CI: 1.50-2.79) times greater than those prescribed benzodiazepines only (85 deaths) and opioids only (64 deaths), respectively, and 1.60 (CI: 1.18-2.15) times greater than nonusers (71 deaths).

Implications:
Concurrently prescribing opioids and benzodiazepines to patients with PTSD was associated with increased all-cause mortality, as compared to prescribing opioids only, benzodiazepines only and to nonusers of these medications. Increased risk of mortality should be considered when considering the potential harms and benefits of co-prescribing opioids and benzodiazepines.

Impacts:
Knowledge gained from this study supports VA efforts to reduce non-recommended prescribing practices, including concurrent prescribing of opioids and benzodiazepines.